Abstract: PA-05-016 Decision Making in Health Behavior Maintenance encourages the formulation of new health behavior models that explain the full spectrum of behavior change, from initiation to long-term maintenance. This proposal will describe the Sentinel Events Model, a new dynamic model which is predicated on the observation that an attempt to quit using an addictive substance is often preceded by the experience of a negative consequence of use, such as a health problem. For some people, this change can be enduring while for others it is only temporary. We seek to better delineate the cognitive and affective mechanisms of action that mediate between a sentinel event and behavior change milestones. Our studies will utilize an innovative measurement strategy: cellular telephone-based ecological momentary assessment (EMA). EMA is the repeated collection of real-time or near-real time data during the routine course of a participant's daily life and controls for the limitations inherent in using aggregated or retrospective measures. Our Research Plan consists of three separate studies. In the Scale Development Study, we will use a mixed qualitative-quantitative method to create and validate measures of the model's constructs. Item Response Theory will be used as an empirically-based method to guide item selection and alternate forms development. In the EMA Refinement Study, we will use the validated scales to guide EMA item construction, and various EMA schedule densities will be examined to determine the frequency that maximizes compliance. Participants will use cellular telephones provided to them for the purpose of the study to complete the EMA. In the Model Testing Study, we will use the refined EMA protocol in a sample of 432 smokers presenting to an emergency department with signs and symptoms that prompt a cardiac evaluation (sentinel health event). The EMA period will involve multiple random prompts each day for the first 28 days, followed by 56 days of end-of- day diaries only. A re-assessment of the model's constructs using the baseline assessment battery will occur at the end of the three month EMA period and again at the six month anniversary of the event. Biological validation of 7-day abstinence will be obtained using saliva cotinine at 1, 3, and 6 months. Our primary objectives center around determining how constructs in our model, in combination with other important predictors of smoking cessation predict: (1) which smokers actually make a decision to quit after a sentinel health event;(2) which smokers actually attempt to quit, defined as 24+ hours without smoking even a puff;(3) which quitters lapse, defined as smoking at least one puff after a quit attempt;and (4) which quitters relapse, defined as smoking on >= seven consecutive days or >= one day on two consecutive weeks. The Sentinel Events Model is distinguished from most existing theories because of its (1) dynamic formulation, (2) inclusion of both cognitive and affective constructs, and (3) emphasis on the role of a discrete event or antecedent in prompting behavior change. Implications for intervention development are discussed. PUBLIC HEALTH RELEVANCE Narrative: The studies outlined herein have important implications for improving public health, because they will elucidate both cognitive and affective factors associated with decision making in the maintenance of substance cessation. Specifically, we have described the model's relevance to the development and enhancement of four categories of interventions: (1) healthcare provider/educator counseling;(2) biomarker feedback interventions;(3) induction of activating affect, both during the event as well as longitudinally;and (4) cell-phone based relapse prevention interventions. The model and approach are innovative, because they involve not only validating a new dynamic model of health behavior, but because of the use of a novel measurement strategy -- cell phone-based ecological momentary assessment.